InformAction Forums

Hi,

When trying to stop loading a faulty or incomplete url the browser still either loads some page or loads a search engine. Only way I found to circumvent this unwelcome feature is to press the stop loading page shortcut several times.

The problem probably lies in the browser doing the loading request TWICE (or more) instead of once. Meaning that just hitting that stop button once will have no effect, because you only stopped one request, leaving the other one still active...

So I tried to disable all add-ons and the issue didn't appear. After some testing the NoScript add-on appears to cause this, regardless of other enabled add-ons.

So my question is: How do I stop this unwelcome behavior?

> a faulty or incomplete url

Explain?
Is this something you've typed in to the URL bar or are you following a bad link?

Examples of such?

> the browser still either loads some page or loads a search engine

May depend on your Location Bar settings & possibly also by your ISP (or DNS service provider).
Default (in SeaMonkey at least) is to perform a web search when urlbar text is not a web location.

>Explain? Examples of such?

When you type part of an URL firefox allows you to scroll through pages you visited in the past. If that down key doesn't register (e.g. pressed to lightly) and you hit enter, firefox tries to still load it as a page, if it can't it redirects you to a search engine... But this is not the problem I have...

The problem is explained after this, so let me copy the relevant part of my first post:

The problem probably lies in the browser doing the loading request TWICE (or more) instead of once. Meaning that just hitting that stop button once will have no effect, because you only stopped one request, leaving the other one still active...

So I tried to disable all add-ons and the issue didn't appear. After some testing the NoScript add-on appears to cause this, regardless of other enabled add-ons.

Can't reproduce it here.

Does it happen on a new profile?

Yes, it happens even with fresh reinstall of NoScript...

Reinstalling NoScript is not the same thing...have you actually created a new profile?

I can actually reproduce this quite reliably in Firefox on all the Windows machines I am using (NoScript ver. 2.6.8.4). I will restate the steps in a possibly clearer manner to illustrate this:

1) Create a new profile, do not install NoScript
2) Type 'z' in the address lookup bar and hit 'Enter', immediately hit 'Esc' to stop page loading
3) No page load, as expected (Search engine lookup, domain guessing do NOT happen - 'Esc' seemingly cancels all of that)
4) Install NoScript
5) Type 'z' in the addres lookup bar and hit 'Enter', immediately hit 'Esc' to stop page loading
6) Page load initially gets cancelled, but ~1 sec later the page loads to whatever is Firefox's default action on entering unrecognised addresses (search engine lookup, domain guessing, whatever is default - the key point is that NOTHING happens when NoScript is not installed i.e. page load is completely stopped)

Of course, there is nothing special about 'z', and any incorrect/unresolvable address (faulty or incomplete in the OP's words) will do instead. Interestingly enough, 'correct' (resolvable) addresses do not trigger this, so I suspect this might have something to do with incorrect address handling scripts or somesuch (I am shooting in the dark here, and have no idea how incorrect addresses are handled, so don't take this too seriously)

Neat, how do you like that!

1) Create a new profile, do not install NoScript
2) Type 'z' in the address lookup bar and hit 'Enter', immediately hit 'Esc' to stop page loading
3) No page load, as expected (Search engine lookup, domain guessing do NOT happen - 'Esc' seemingly cancels all of that)
4) Install NoScript
5) Type 'z' in the addres lookup bar and hit 'Enter', immediately hit 'Esc' to stop page loading
6) Page load initially gets cancelled, but ~1 sec later the page loads...

It's true. Does happen.

therube wrote: It's true. Does happen.

Ummm, does that mean it's expected behaviour? Fixable? Fix-worthy? The only reason I pointed this out was that it's been bugging me, and Thrawn seemed not to be able to reproduce it. There was also a thread from oabout 4 years ago which died out after another Not Reproducible post, and these are the only two threads I can find with any information on this at all.

Certainly able to reproduce (now with your STR).
Given that it works differently with NoScript installed, I would say, Bug.
All the rest, that'is up to the powers to be.

therube wrote:Certainly able to reproduce (now with your STR).
Given that it works differently with NoScript installed, I would say, Bug.
All the rest, that'is up to the powers to be.

Bug? yes.
Fixable? probably, but it's related to a quite complex and fragile machinery (request redirection).
Also, it happens only for first-hit domains, so it doesn't seem a big impairment.
Thus, low-priority bug to be fixed.